On Dec 8, 2003, at 10:25 AM, Philippe Verdy wrote:
At least, IDC would really speed up the work of unification of various
repertoire sources and help avoid duplicates. It seems that, even if IDC
allows several compositions, a sort of canonical ideographic decomposition
(not a NFD or NFKD decomposition!) could be created to match these
characters with a precise 2D composition model based on IDC's.
There are too many ambiguities and other problems that make this wildly impractical.
Such database would also help the creation of ideographic fonts, by allowing
to compose rapidly a first version of a complete Han font with very few
glyphs for radicals or strokes (after this automatic step, it could be
compared to carefully designed glyphs, to author the font with careful
adjustments such as kerning, partial overlays, or small variations of stroke
widths/length/angle.
This would mean lower costs to develop a functional Han font.
Actually, IDSs are ill-suited for making fonts, since they under-specify data. The CDL discussed at the last UTC is far, far better, and other systems are actually used by font designers.
======== John H. Jenkins [EMAIL PROTECTED] [EMAIL PROTECTED] http://homepage..mac.com/jhjenkins/
